See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266370390

White cane sugar production

Article  in  International Sugar Journal · January 2009

CITATIONS READS
3 16,565

1 author:

Peter Rein
University of Cape Town
56 PUBLICATIONS   396 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Peter Rein on 13 February 2015.

The user has requested enhancement of the downloaded file.

 WHITE CANE SUGAR PRODUCTION
P.W. Rein
Professor Emeritus, Louisiana State University
Consultant

Abstract

The options and prospects for the direct production of white sugar in a cane sugar mill are
reviewed. Plantation white sugar has been made for many years in some countries but cannot
meet the requirements for refined sugars. These low colour sugars are now a popular quality raw
sugar for the increasing number of destination refineries. Attainment of refined sugar standards
needs the application of some or all of the new separation technologies such as membrane
separation, the use of adsorbents, and chromatographic separation. Only one of the new processes
being proposed is close to commercialization, namely the WSM process developed by Tongaat
Hulett Sugar. A new process for the production of white sugar in a cane sugar mill developed at
the Audubon Sugar Institute is described in some detail. It does not require membrane separation
and involves adsorption of colour and other impurities using granular activated carbon and ion
exchange resins. Chemical regeneration of the carbon is utilized, which enhances the
attractiveness of the process. Preliminary information is given on a possible full-scale installation
in a raw sugar mill.

Introduction

Some significant trends are evident in the world sugar market. In the 1980s the proportion of
white sugar traded on the world markets, both beet and cane showed a clear increase relative to
the raw sugar (Clarke 1986). More recently, the trend has changed, since a number of new large
refineries have been built and white sugar exports from producing countries, particularly from the
EU, have reduced. Another trend evident is the swing to VHP or better raw sugar supply to
refineries, and in particular an increase in deliveries of very low colour raws exemplified by
Brazilian “cristal”.

Todd (1997) has shown that the average cost of production of refining in attached refineries is
about 55 % of the cost in stand-alone refineries. In order to compete, autonomous refineries have
to be larger to achieve economies of scale and lower unit production costs. A further trend
evident in the US has been the increase in vertical integration from field to white sugar output.
All beet sugar factories in the US are now owned by the growers, and many large raw cane sugar
producers have invested in refining, both white-end refineries and in existing or new refinery
capacity.

The first new stand-alone refinery to be build for 25 years was constructed in 1990 in Dubai.
Since that time, over a dozen new refineries have been built and existing refineries have been
expanded. Currently a number of new refineries are being built or are planned in areas such as
Bangladesh, Egypt, Syria, India, Louisiana and Israel among others.

1
The relationship between white and raw exports has reversed, with raw sales now increasing
relative to whites. Freight differentials, between the costs of shipping raw sugar in bulk and
bagged white sugar, are partially responsible, particularly with larger vessels now shipping bulk
raw sugar. In many countries too the tariff protection for import of white sugar is higher than for
raw sugar.

Another factor promoting the development of new destination refineries is the considerably
enhanced raw sugar quality available. Sugar quality has a major impact on refining costs and an
interesting example of a refinery working together with a raw supplier was reported by Hikmat
and Oliveira (2006). Improvements in sugar quality to achieve a polarization of over 99.6 °Z
resulted. This study validated their statement: “Reducing the impurities in raw sugar is a
governing factor in improving performance of the sugar refinery and can be achieved easily and
economically in the sugar mill”. This effort has resulted in the Dubai refinery being among the
lowest cost refiners in the world.

The threats to the refiners are a reduction in the white premium, a new higher sustained level of
energy costs and the prospect of direct white sugar production technology developed to a feasible
stage (e.g. Fechter et al. 2001; Rossiter et al. 2002; Kochergin et al 2001; Oliverio and Boscariol
2006; Rein et al. 2006).

The raw sugar produced in a cane sugar mill is normally sent to a refinery, to be converted from a
light brown raw sugar with a purity of around 98.5 % to an almost pure (>99.9 %) white sugar for
sale to consumers. A flow diagram showing the unit operations in a raw sugar mill is shown in
Figure 1.

Bagasse

Cane
Extraction Heating Clarification

Raw sugar
Evaporation Crystallization Centrifugation

Molasses

Figure 1: Raw sugar mill flow sheet

The refining process incorporates colour removal steps, filtration and crystallization. The latter is
a particularly efficient purification process, but is expensive in terms of energy and equipment.

Mill white sugar has been produced at raw sugar mills for many years without refining or
recrystallization. This has generally involved the use of juice and/or syrup sulphitation, producing
a sugar with a colour of around 100 to 200 IU. Nowadays standards for white sugar are
considerably higher, requiring a colour below 45 IU for EU No 2 or even 22.5 IU for grade 1
white sugar.

2
More recently many attempts have been made to produce white sugar directly in a raw sugar mill
without melting (dissolving) and re-crystallizing through the use of new and innovative
separation technologies. The prize is the significant additional value of the product, with a white
premium consistently over $60 /tonne (Todd 1997), as well as lower energy costs, because the
mill uses sugarcane bagasse as a fuel.

Plantation white sugar

Indian sulphitation process

Juice sulphitation has a remarkable decolourizing effect in some industries, particularly India,
where it is widely practised. Sulphur is burnt in a sulphur burner under carefully controlled
conditions to produce SO2 while avoiding sublimation and the production of SO3. Contact of the
gas with juice pre-heated to about 70 °C takes place in a sulphitation tower or a series of tanks.
The precipitate formed due to the addition of lime and SO2 separates in juice clarification. This
process is generally used together with syrup sulphitation in flotation clarification. Sulphur
consumption is 250 to 500 mg sulphur/kg cane if only juice sulphitation is practiced. If both juice
and syrup are sulphited, usage can be as high as 900 mg/kg cane. The advantages of sulphitation
are:

• The mud settles more rapidly (hence improved capacity of clarifiers).

• The massecuites are less viscous and boil faster.
• There is a marked improvement in the colour of the sugar – there appears to be a
significantly reduced transfer of colour from the mother liquor into the crystal.
• There is a slight gain in the capacity of centrifugals.
However there are also some disadvantages:
• There are heavier fouling deposits in heaters. These may largely be avoided by sulphiting
hot, but this involves an increase in the heating surface required.
• An increase in chemical and maintenance costs (in sulphitation tanks, circulation pump,
sulphur, corrosion of vessels and pipes, etc.).
• Heavier use of lime in the process and in the recirculating cooling water.
• The residual sulphur in the sugar may exceed the specification demanded by some
customers.

Brazilian process
Juice sulphitation is common using sulphitation towers or juice ejectors, using 150 to 300 mg
sulphur/kg cane. However it is more common to see syrup clarification carried out using
phosphatation, with the addition of phosphoric acid at 50 to 150 mg/kg sugar, and lime. This
helps to get the SO2 content of the sugar down below the 10 mg/kg specification.

The Brazilian production of low colour sugar is materially assisted by the simultaneous
production of ethanol. This enables to mill to choose the higher purity streams (e.g. primary
juice) to be used for sugar production and the lower purity streams go to ethanol production; in
addition there is no recycle of low purity streams back to sugar production. In addition, a lower
pH profile in the mill can be employed to minimize colour formation; inversion of sucrose is not
an issue since the reducing sugars formed are still utilized in ethanol fermentation.

3
Blanco Directo sugar
Tate and Lyle introduced the Blanco Directo process, which involves filtrate flotation
clarification and syrup clarification, together with phosphoric acid and lime addition, very similar
to the refinery phosphatation process (Bennett and Ross 1988). A comparison by Trott (1988) of
the qualities of refined sugar with direct mill white sugars is shown in Table 1.

Refined Blanco Directo Mill white

Polarization 99.9 99.8 99.6
Colour (IU) 10-80 100-200 200-250
Turbidity (ICUMSA) 10-30 20-50 100-500
Ash (g/100g) 0.01-0.04 0.05 0.1
Invert sugar (g/100g) 0.01-0.04 0.02 0.1
SO2 (mg/kg) - 1-5 20-50

Table 1. Comparison of white sugar specifications

The benefits of syrup clarification are widely realized and it is becoming a more popular process,
not only in Brazil, particularly when a higher quality sugar is required for subsequent refining. It
can remove about 85 % of the turbidity, gives a small ash and colour improvement and reduces
massecuite viscosity. Its effect on sugar quality is generally greater than the improvements
measured in syrup quality suggest (Steindl and Doherty 2005), largely because high molecular
mass impurities are removed. Filtrate clarification is still successfully used in some mills, but has
been found to be difficult to operate in others and has been abandoned where tried in South
African mills.

Options for white sugar production in a raw mill

The Dedini DRD process

Dedini have taken the Blanco Directo approach a stage further by introducing ion exchange. The
process includes juice sulphitation, syrup clarification/phosphatation, deep bed filtration and then
three ion exchange stages for softening (demineralization), decolorization and polish decolorizing
(Oliverio et al. 2007). This process has been run as a pilot plant in Brazilian mills for two years.
It is claimed that on average 52 % decolorization is achieved with ion exchange giving a syrup of
4200 IU, regenerating the columns with NaCl and NaOH. This syrup boiled in a small pan gave
sugar meeting EU No. 2 colour specifications, indicating a very low colour transfer, intermediate
between that obtained in the beet and cane industries.

Dedini are promoting this process for new installations or as a retrofit into an existing sugar mill.

4
Membrane separation
Developments in membrane technology have been such that they can cope with the demanding
requirements in a sugar mill, namely the ability to handle high temperatures and varying
suspended solids characteristics including abrasive particles, while still achieving good
separations at high flux rates and reasonable cost. Ceramic, stainless steel/ceramic and polymeric
membranes are now available which can meet these requirements.

Although great strides have been made in terms of both cost and technology, the use of
membranes for decolourization of juice is limited. It is not feasible to apply membranes to raw
juice, but they have been tried in many situations using clarified cane juice. This has to be done at
high temperature around 90 °C to ensure that no microbiological loss or fouling occurs. Steindl
(2001) surveyed reported results which show an average colour removal from clear juice of 14 %
using ultrafiltration, based on different membrane molecular weight cutoff values down to 50
kDa. There is some evidence to suggest that the improvement in colour of sugar may be slightly
higher than the reduction in juice colour implies. Nevertheless it does not appear that membranes
alone can enable the direct production of white sugar. In addition, the recovery of the sugar
remaining in the retentate is costly.

No improvement in recovery of sugar results because there is only a marginal improvement in

purity after membrane filtration (Kochergin et al. 2000). It appears that in future membranes may
only find application in raw sugar mills in combination with other processes which add value.

The new Applexion Process (NAP) was developed in Hawaii, using membrane separation and
decolorization (Kwok 1996). This process operated for some time in a raw sugar mill in Hawaii
but has been discontinued. It also does not have the ability to remove ash to achieve refined sugar
standards.

ARi Chromatographic separation technology

Chromatographic separation has been used for many years to recover sugar from beet molasses
and is a well-developed process. It has not found much use in the cane industry, because of the
lower sucrose content, the suspended solids content and the high calcium content in cane
molasses which interferes with the process. The advantage of chromatographic separation lies in
the fact that a high purity sucrose stream can be isolated, from which refined sugar can be
produced, but in addition other products of higher value can be recovered.

Direct white sugar production involves either the removal of the sugar from the juice or the
removal of colour and ash from the juice. This process developed by ARi (Amalgamated
Research Inc.), which involves the use of chromatographic separation, is an example of the
former. It is preceded by membrane filtration.

Use in a cane sugar mill requires the juice to be softened before chromatography can be applied.
This is an expensive process, but would have the added advantage of eliminating evaporator
scaling. The cost of the process could however be reduced by the use of fractal distributors
(Kearney et al. 2001). Until recently, it was thought that softening would have to be preceded by
membrane separation; however trials have shown that possibly this may not be necessary.

5
A block diagram of this process is shown in Figure 2. After softening, the juice is evaporated,
before going to the first stage of chromatographic separation where most of the invert sugars and
low molecular mass components, mainly ash, are removed. A second stage is necessary to
remove mostly inorganic non-sugars, before the sugar rich stream is concentrated prior to
crystallization. In addition to being able to crystallize white sugar directly, the loss of sucrose in
molasses is much reduced; an increase in overall recovery of 6.8 % is claimed.

Bagasse

Cane
Extraction Heating Clarification

Softening Ultrafiltration

White sugar
Evaporation Crystallization Centrifugation

Molasses

Chrom. Sep. 1 Evaporation

Chrom. Sep. 2

Figure 2. Block diagram of ARi chromatographic separation process.

The disadvantage of this process is the very large inventory of resin required (Fechter et al. 2001)
and high pumping costs. In addition the dilution that takes place during chromatography leads to
a very high evaporation load. The results of extensive trials in a raw sugar mill are reported by
Kochergin et al. (2001).

The Tongaat Hulett WSM process

It has been found that in the process of demineralizing juice using ion exchange resins a
significant degree of decolourization is achieved. This is the basis for the WSM process for direct
production of white sugar (Fechter et al. 2001, Rossiter et al. 2002, Jensen 2007). It incorporates
membrane filtration ahead of ion exchange demineralization and also ion exchange
decolourization using the same resins as used in a refinery, to achieve EU No 2 white sugar
specifications.

6
The process does have some other significant advantages, namely (Rossiter et al. 2002):
• an increase in sugar yield
• a large improvement in sugar quality
• the production of a high grade molasses for by-product use
• the removal of species that foul the evaporators
• higher heat transfer coefficients in pans and evaporators

The process is represented schematically in Figure 3. The juice is concentrated to 25 Brix

because the maximum dissolved solids flux is achieved on membrane separation at this
concentration. Cooling of the juice is necessary before cation exchange treatment because
significant inversion could otherwise occur at the low pH values experienced in cation exchange.
Some heat recovery is achieved in reheating the treated stream. The ion exchange processes are
carried out in a simulated moving bed system, with 36 cells for each of the three ion exchange
steps (Jensen 2007), which reduces the resin inventory, regeneration chemical usage and water
consumption. Disposal of the membrane retentate becomes a problem if the process is not
operated as a slip stream in a mill.

Bagasse

Cane
Extraction Heating Clarification

White sugar
Evaporation Crystallization Centrifugation

Molasses

Ultrafiltration Heating/cooling

Cation ionex

Anion ionex Decolorization

Figure 3. The WSM process for direct white sugar production incorporated in a raw sugar mill.

The process has been running as a slip stream process at Felixton in South Africa for the last 3
years, designed to process 15 % of the mill’s clarified juice through the WSM process. During
the 2006 season over 5000 t white sugar with an average colour of 37 IU and meeting all EU No.
2 specifications was produced. Various problems largely associated with peripheral equipment
have been experienced and the process is not yet in commercial production.

7
The process is characterized by high chemical requirements for ion exchange demineralization.
The use of suitable regenerants can produce effluents rich in potassium and nitrogen, which could
overcome a disposal problem if used to fertilize the fields.

Development of a new direct white sugar production process

The objective of the work done at Audubon Sugar Institute was to investigate options for
producing white sugar directly without the use of ultrafiltration. Mill clarified juice was treated
with a combination of granular activated carbon (GAC) column decolourization and ion
exchange columns to remove ash and additional colour. The use of oxidants in a combination
with the treatment was also investigated.

The process investigated was expected to achieve the same advantages as the WSM process
given above. In addition it eliminates the need to clean first effect evaporators.

Activated carbon is commonly used for decolourization in sugar refineries, but has not been
applied in treating cane juice. Because of the low viscosity of cane sugar relative to refinery
liquor, higher flow rates are possible. The carbon columns are also capable of handling some
turbidity in the juice without serious fouling.

Oxidants have been used to obtain additional decolourization in refining when circumstances
demand it. They destroy colorants by attacking susceptible functional groups without attacking
sucrose and have the advantage that they have no adverse environmental effects.

Bento (2004) reported that hydrogen peroxide added before resins has a number of beneficial
effects. Conjugated double bonds are broken and carboxylic acids are formed. This decreases the
molecular weight of the colorants and increases their propensity to be removed by ion exchange
resins. Dosage rates of 500 mg/kg DS resulted in longer cycles and higher rates of colour
removal. As a consequence, chemical and effluent disposal costs are reduced.

Initial laboratory investigations

Pure sucrose is colourless, but may appear coloured because of the inclusion of small amounts of
coloured material in the sugar. Colour is the generic term used to cover a wide range of
components which contribute to the colour of sugar. Most of these compounds are complex and
not easy to quantify and colour is measured as the total effect of all colorants on light absorbance
at 420 nm. This is not altogether satisfactory, because different colorant components behave
differently in the various processing and decolourizing operations.

Work done on modelling the removal of colour in the resin columns of the WSM process has
been reported by Broadhurst and Rein (2003). Attempts were made to characterize colorants on
the basis of their molecular mass, using GPC measurements. Suitable deconvolution of the
retention time peaks was used to yield quantitative data on various molecular mass pseudo-
components (Broadhurst and Rein 2003a).

8
This work was done on ultrafiltered juice samples. Subsequent work was aimed at modifications
to the process to obviate the use of membranes, because the cost of membrane separation
militates against the economic viability of such processes, in spite of recent advances in the
technology.

Granular activated carbon (GAC) was substituted as an initial juice treatment step. It is less
prone to fouling than resins, and is a useful decolourizing process in its own right. A high juice
flow rate through the GAC column of 15 BV/h was used. It was still necessary to cool the juice
after the carbon columns to below 10 °C to prevent inversion in the ion exchange columns. A
comparison of the juice colour obtained using the carbon column and ultrafiltration prior to ion
exchange showed that using GAC, a juice colour of 2000 ICUMSA units (IU) was possible even
after repeated regenerations, about half of that using membrane separation before ion exchange
(Ellis 2004). Juice feed colour averaged about 12 000 IU. Regeneration of the carbon was
undertaken in a laboratory furnace at 600 °C.

The effect of adding hydrogen peroxide was investigated for this process in conjunction with
carbon treatment. The results of some laboratory batch tests are shown in Figure 4. This indicated
that additional colour removal could be achieved when peroxide treatment was combined with
GAC decolourization. The tests showed that little colour was removed by peroxide on its own.

12000

10000

8000
Color IU

6000

4000

2000

0
0 20 40 60 80 100
Time (min)

Feed GAC
H2O2 (1000 mg/kg) GAC + 200 mg/kg DS
GAC + 500 mg/kg DS GAC + 1000 mg/kg DS

Figure 4. Batch tests on colour removed from juice by combinations of carbon and peroxide
treatment at 80 °C.

It was found that the modified process using a carbon column could be substantially improved by
the addition of hydrogen peroxide to the juice before the carbon columns. Because the colour of
the juice is much higher than the colour in refinery streams, a dosage rate of 1000 mg/kg DS was
used ahead of the carbon columns. A reaction time of 30 minutes at a temperature of 85 °C was
used.

9
Another important aspect of the resin performance is the deashing ability of the resin. The
combination of carbon and peroxide also leads to a far superior overall ash removal.

Factory trials.
Trials have been run over three seasons in Louisiana sugar mills. More detail is given elsewhere
(Rein et al. 2006). The first trial investigated only the performance of GAC in treating clarified
juice, pre-treated with hydrogen peroxide over a 10 week period. The carbon was chemically
regenerated once/week, using a combination of washes of HCl, NaOH, ethanol and a special
additive (Bento 2006). On average the juice colour was reduced from 12 000 to about 3 000 IU.

The system proved to be fairly robust, handling periodic entrainment from the clarifier, but in
spite of this still achieving substantial decolourization. The first step in regeneration involved
backflow at a high rate to remove entrapped mud from the column.

Ash removal was minimal, as expected. The pH generally dropped across the column, as the acid
wash appeared to take out the magnesite in the carbon which would otherwise maintain a neutral
outlet pH. Variations in temperature over the test period appeared to have little effect. A more
important factor was the mud carry-over from the clarifier. On two occasions after heavy rains,
the column had to be back-washed with water to restore the required juice flow rate.

A second trial was conducted at St. James mill during the 2005 season. Some changes were
made to the test program. Apart from eliminating experimental problems experienced the
previous year, the trial was modified to assess a different ion exchange treatment. Instead of
employing strong acid cation and weak base anion columns after the carbon, as was practiced in
the laboratory trials, a mixed bed of softening and decolourizing resin was employed. This
mixture did not change the pH of the juice, and so dispensed with the need for cooling the juice
to a low temperature to avoid inversion.

A small GAC guard column was used to protect the second column from suspended solids
carried over with the juice. Before the second column, juice was contacted with hydrogen
peroxide. After passage through the GAC columns, juice was treated with cationic and anionic
resins in a mixed bed column, to remove hardness in solution and to reduce juice colour. The
installation worked continuously for two months with clarified juice at a flow rate of 1 BV/h (on
total carbon). Average data are shown in Table 1. A reduction of 81.5 % in juice colour was
obtained over the test period (75.6 % on carbon and 24.1 % on resin).

Table 1: Composition of clarified juice before and after treatment

Colour Turbidity Ash

Brix pH
(IU) (NTU) (g/100 g DS)
Clarified juice 14.3 9 070 212 7.7 3.5
Juice after 1st GAC column 14.3 6 870 191 5.9 3.7
Juice after 2nd GAC column 14.1 2 220 187 5.6 3.9
Juice after ion exchange 13.9 1 680 165 6.0 4.0
Reduction (%) 81.5 22.2

10
A turbidity decrease of 22.2 % (Table 1) and hardness (calcium and magnesium) reduction of
88.5 % was observed (Table 2). An increase in lactic acid through the columns was also
registered at times, because it was not possible to keep the temperatures consistently above 80 °C
in the pilot plant.

Table 2. Composition of juice

Apparent purity Ca + Mg content

(%) (mg/kg DS)
Clarified juice 87.4 4240
Juice after 1st GAC column 88.6 4460
Juice after 2nd GAC column 88.2 4350
Juice after resin column 89.0 489
Reduction 88.5 %

Resin cycles varied between three and four days. After each cycle, the resins were washed with
condensate water followed by regeneration with NaCl and NaOH. The performance of the ion
exchange resins was affected by contamination of the “condensate” stream with low quality
water, which adversely affected regeneration. Nonetheless it appears that this option will not
reduce the ash content of juice sufficiently to be able to achieve EU No 2 sugar ash
specifications; an ash reduction of at least 75 % is necessary.

After each five day cycle, carbon was washed and regenerated chemically. Colour loading was
high, well over 200 t DS·IU/L resin per cycle.

Some analyses done on four samples during one day of the trial gave the values of colour and
Indicator Value shown in Figure 5. In this case the Indicator Value is the ratio of colour
measurements at 8 and 5 pH instead of the usual values of 9 and 4 pH. This shows a steady
decrease in Indicator Value, suggesting that a substantial part of the natural colorants (phenolics
and flavanoids) is removed. Bento (2006) evaluated the propensity of other juice colorant types to
be adsorbed on GAC.
5.0 8000

7000
4.5
Indicator Value (pH 8/pH 5)

6000
4.0
5000
Color (IU)

3.5 4000

3000
3.0
2000
2.5
1000

2.0 0
1 2 3 4
Clarified Juice After GAC 1 After GAC 2 After Resin
Sample
Figure 5. Measurements of colour and Indicator Value
11
A third trial was run at Raceland mill in 2006, with demineralization following the initial GAC
treatment. The juice was refrigerated before resin demineralization. The trial was seriously
affected by equipment problems and the plant was unable to run the full process continually for
long enough to establish some of the required design parameters required. However it did appear
that more than 50 % decolourization could be achieved routinely with GAC and 80 %
decolourization over the resin columns when they were properly regenerated. The trial also
showed that the resin performance was not affected by resin fouling. Again the short season
proved to be a problem in getting pilot plant equipment to work reliably in a short period of time.

Application to a raw sugar mill.

The flow sheet for a raw sugar mill using this process is shown in Figure 6. Both the process and
the regeneration are the subjects of patents (Rein et al. 2007; Bento et al. 2008). A syrup clarifier
is incorporated to ensure that the suspended solids content is reduced sufficiently for the high
quality sugar required. It goes without saying that juice clarification must be well-controlled to
ensure that the columns are not subject to carry-over from the juice clarifiers at any time.

Bagasse

Cane
Extraction Heating Clarification

White sugar
Syrup Crystallization Centrifugation
Clarification
Molasses

Evaporation

Carbon Heating/cooling

Cation ionex Anion ionex

Figure 6. Alternative process for direct white sugar manufacture in a raw sugar mill.

A number of options exist for the configuration of the GAC system. Instead of a guard column, a
pulsed bed GAC system would on a regular basis discharge the carbon layer which could get
fouled by carry-over from the clarifiers. A substantial advantage of the chemical regeneration
system however is the ability to configure the GAC system in the form of a simulated moving
bed. The advantages, as with ion exchange, could be substantial (Hubbard and Dalgleish 1996),
in particular reducing the cost of regeneration and reducing the amount of sweet water produced.

12
Chemical regeneration of GAC seems viable, at least over a period of 10 weeks. Beyond this time
it may be necessary to introduce thermal regeneration of part of the GAC to keep the process
operating efficiently. Transport of carbon to an outside facility would be the best option if
transport costs allow it. In thermal regeneration, something between 4 % - 10 % of the GAC is
lost in the regeneration process. Thus after 10 regenerations a significant amount has already
been lost in the process of thermal regeneration. This is avoided in chemical regeneration.

The use of ethanol in regenerating GAC is particularly suitable in a sugar mill which also
produces ethanol, with the following advantages:
• The spent regenerant can be sent back to the distillation plant for a total recovery of the
ethanol.
• A periodic ethanol wash also improves the decolourization capacity of the SAC column
(Broadhurst and Rein 2003), which could be recovered.
• Inversion to reducing sugars is not important, eliminating the need for refrigeration before
ion exchange

The cost of regenerants for the process presented here is estimated to be roughly $0.50 for a
colour loading of 1 t DS·IU/L of carbon per cycle. In the Louisiana context, this translates to 0.40
US cents/kg sugar for regenerating GAC. The cost of regenerating the ion exchange columns is
1.67 US cents /kg sugar, giving a total of about 2 US cents /kg. This can be reduced to < 1.50 US
cents with some reuse of the HCl and NaOH.

The disposal of ion exchange effluent is a problem or perhaps an opportunity depending on the
particular mill’s circumstances, as in the case of the WSM process. The inorganic material
removed is that present in the cane, and is particularly high in potassium, which has considerable
fertilizer value. Regeneration of the WBA columns with ammonium hydroxide can significantly
boost the fertilizer value of this regenerant (Rossiter et al. 2002). The use of the regenerant as a
fertilizer could be the best solution if it could be integrated into the combined operation of
growing and milling without any detrimental effects on the environment.

Additional power and steam requirements are expected to be modest, in contrast to processes
involving membranes which incur high power costs.

Conclusions

Considerable attention is being given to the direct production of white sugar in a cane sugar mill.
Improvements in separation technologies are making some approaches look viable and will
probably be used to produce better quality sugar if not sugar of refined quality. Combined with
the advantages of higher sugar recovery, better molasses quality and the elimination of
evaporator scaling, prospects for direct white sugar production are good. The step from small
scale proving of a process to full scale viability is a difficult one to take and further work is still
necessary to establish optimal design criteria and financial viability.

References

13
Bennett M.C; Ross B.G. (1988): Blanco Directo production, in Raw Sugar Quality and White
Sugar Quality. Ed Clarke M.A. Verlag Dr A. Bartens, Berlin, 89-100.

Bento L.S.M. (2004): Decolourization of sugar solutions with oxidants and ion exchange resins.
Proc. Sug. Ind. Technol. 63, 265-269.

Bento L.S.M. (2006): Activated carbons: adsorption of sugar colorants and chemical regeneration.
Proc. Sug. Ind. Technol. 65.

Broadhurst H.A.; Rein P.W. (2003): Modeling adsorption of cane-sugar solution colorant in
packed-bed ion exchangers. AIChE Journal. 49, 2519-2532.

Broadhurst H.A.; Rein P.W. (2003a): Deconvolution of GPC chromatograms of sugar solutions.
Zuckerindustrie 127, 3, 96-99.

Clarke M.A. (1986): Cane sugar production systems: current and future. Proc. S. Afr. Sugar
Technol. Ass. 60, 1-4.

Ellis B.M. (2004): Modeling of Cane Sugar Colorant Removal in Packed-Bed Ion Exchange
Columns and an Investigation into Pretreatment methods. MS Thesis, Chemical Engineering,
Louisiana State University.

Fechter W.; Rajh S.; Reimann M.; Ahmed F.E.; Jensen C.R.; Schorn P.M.; Walthew D.C. (2001):
Direct production of white sugar and high-grade molasses by applying membrane and ion-
exchange technology in a cane sugar mill. Proc. Int. Soc. Sugar Cane Technol. 24, 100-107.

Hikmat O.; Oliveira D.T. (2006): Quality improvement of raw sugar - working with your
supplier. Proc Sugar Proc. Res. Conf. 72-79.

Hubbard G.M.; Dalgleish G.B. (1996): Decolourization of carbonatated liquor using the ISEP
principle of continuous ion exchange. Sugar Ind. Technol. 55, 23-42.

Jensen C.R.C. (2007): Direct white sugar manufacture in the cane sugar industry via membrane
filtration and continuous ion-exchange demineralization. Zuckerind. 132, 6, 446-452.

Kearney M; Velasquez L.; Peterson K.; Mumm M.; Jacob W. (2001): The fractal softener. Proc.
Amer. Soc. Sugar Beet Technol. 31.

Kochergin V. (1999): New separation technologies. Sugar Proc. Res. Inst. Symp. on Advanced
Technology for Raw Sugar and Cane and Beet Refined Sugar Production. 40-48.

Kochergin V.: Kearney M.; Jacob W.; Velasquez L.; Alvarez J.; Baez-Smith C. (2000):
Chromatographic desugarization of syrups in cane mills. Proc. S. Afr. Sugar Technol. Ass. 74, 62
-69.

Kochergin V.; Kearney M.; Alvarez J. (2001): Direct production of white sugar in cane mills:
technical and economic aspects. Proc. Int. Soc. Sugar Cane Technol. 24, 1, 108-111.

14
 Kwok R. J. (1996): Ultrafiltration / softening of clarified sugarcane juice using the new
Applexion process. Sugar Proc.Res.Inst. Workshop on Separation Processes in the Sugar
Industry, 87.

Oliverio J.L.; Boscariol F.C.; Filho H.G. (2007): DRD – Dedini Refinado Direto (Dedini Direct
Refined): refined sugar without a refinery. Proc. Int. Soc. Sugar Cane Technol. 26, 1612-1621.

Rein P.W.; Bento L.S.M.; Cortes R. (2006): The direct production of white sugar in a cane sugar
mill. Proc Sugar Proc. Res. Conf. 112-127.

Rein P.W.; Bento L.S.M.; Ellis B.M. (2007): Direct production of white sugar from sugarcane
juice or sugar beet juice. US Patent No. 7226511.

Bento L.S.M.; Rein P.W. (2008): Chemical regeneration of activated carbon. US Patent Pending.

Rossiter G.; Jensen C.; Fechter W. (2002): White sugar from cane at the factory. The impact of
WSM. Proc. Sugar Proc. Res. Conf. 162-177.

Steindl R.J. (2001): Membrane filtration technology in the cane sugar industry. Proc. Int. Soc.
Sugar Cane Technol. 24, 3-10.

Steindl R.J; Doherty W.O.S. (2005): Syrup clarification for plantation white sugar to meet new
quality standards. Int. Sugar J. 107, 1282, 581-589.

Todd M. (1997): Prospects for the world white sugar market. Proc. Sugar Ind. Technol. 56, 1-14.

Trott R.R. (1988): Clarification and decolourization processes, in Chemistry and Processing of
Sugarbeet and Sugarcane. Eds Clarke M.A. and Godshall M.A. Elsevier, Amsterdam. 265-291.

BSST_WhiteSugar_2008
31 March 2008

15

View publication stats